View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by University of Essex Research Repository

International Journal of Law and Psychiatry 64 (2019) 60–70

Contents lists available at ScienceDirect

International Journal of Law and Psychiatry

journal homepage: www.elsevier.com/locate/ijlawpsy

Objectivity in science and law: A shared rescue strategy T ⁎ Matthew Burcha, , Katherine Furmanb

a University of Essex, School of Philosophy and Art History, Wivenhoe Park, Colchester CO4 3SQ, United Kingdom b University College Cork, Department of Philosophy, College Road, Cork, Ireland

ARTICLE INFO ABSTRACT

Keywords: The ideal of objectivity is in crisis in science and the law, and yet it continues to do important work in both Objectivity practices. This article describes that crisis and develops a shared rescue strategy for objectivity in both domains. Science In a recent article, Inkeri Koskinen (2018) attempts to bring unity to the fragmented discourse on objectivity in Law the philosophy of science with a risk account of objectivity. To put it simply, she argues that we call practi- Psychiatry tioners, processes, and products of science objective when they identify and manage certain important epistemic Epistemic risk risks. We endorse this view and attempt to tailor Koskinen's strategy to the problem of objectivity in the legal Phronetic risk context. To do so, we develop a novel notion of phronetic risk, and argue that we call practitioners, processes, and products of law objective when they identify and manage certain important epistemic and/or phronetic risks. Our attempt to rescue objectivity is especially important for work at the intersection of law and psychiatry. For that intersection represents a place where skeptical worries about objectivity in science and law work in tandem to pose serious critical challenges to contemporary practice; and our rescue strategy represents a pro- mising way to negotiate those challenges.

1. Introduction specialties” (Pies, 2007, p. 20). Concerns over objectivity in psychiatry thus pose two main questions: 1) is there such a thing as objective Objectivity is a central concern for many of the issues at stake in this psychiatric disorder out there in the world; 2) if objective psychiatric journal; the intersection of law and psychiatry is a place where skeptical disorder exists, can we objectively identify and measure it? worries about objectivity in science and law work in tandem to pose A similar back and forth plays out in the relevant legal discourse. serious critical challenges to contemporary practice. For instance, The Committee on the Rights of Persons with Disabilities echoes the concerns about the scientific objectivity of psychiatry abound in the critics of psychiatry in its General Comment No 1 (UN Committee on the literature on mental disorders. Skeptics treat the term “mental disorder” Rights of Persons with Disabilities, 2014), when they claim, “Mental as a purely evaluative term designed to justify medical intervention capacity is not, as is commonly presented, an objective, scientific and (Szasz, 1974); skeptics and non-skeptics alike point out that the diag- naturally occurring phenomenon (par. 14).” Moreover, many disability nosis of mental disorders sometimes serves as a social tool of oppression rights advocates argue that mental health law fails to be objective in that stigmatizes perfectly healthy but aneurotypical people (Eysenck, that it reflects blatant bias against persons with intellectual andpsy- Wakefield, and Friedman, 1983); and non-skeptics who hold that chosocial disabilities (Harpur, 2009), arguing that the law takes cover mental disorders have an underlying physiological basis still insist that behind the rhetoric of objectivity, while it stigmatizes and imposes the concept has an ineliminable value component (Wakefield, 1992). “ableist” norms on persons with intellectual and psychosocial dis- Moreover, this evaluative dimension of the concept of mental disorder abilities. But this skeptical viewpoint does not predominate. The Eur- drives worries that the field of psychiatry is just not as objective as opean Court of Human Rights, for instance, specifies that the lawful other areas of medicine. On the other side of this debate, however, deprivation of a person's liberty on the grounds that she is of “unsound defenders of psychiatry's objectivity point out that it operates within mind” must be based on “objective medical expertise” and “objective the scientific medical model (Shah & Mountain, 2007), and they argue medical evidence” (European Court of Human Rights, 2012 par. 103, that if we take interrater reliability as a measure of objectivity, “psy- emphasis added). Far from an idle theoretical question, then, objec- chiatric diagnosis is often as objective as that in most other medical tivity is a central concern for contemporary public policy challenges at

⁎ Corresponding author at: London School of Economics and Political Science, Centre for Philosophy of Natural and Social Science, Houghton St., London WC2 2AE, UK. E-mail addresses: [email protected] (M. Burch), [email protected] (K. Furman).

https://doi.org/10.1016/j.ijlp.2019.02.004 Received 17 October 2018; Received in revised form 12 February 2019; Accepted 20 February 2019 0160-2527/ © 2019 Published by Elsevier Ltd. M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 the intersection of psychiatry and the law. of objectivity in science and apply it to the law. Koskinen argues that we Unfortunately, the concept of objectivity is in a crisis in science and call practitioners, processes, and products of science objective – and so the law. In both domains, the term is multiply ambiguous, referring to we trust them and believe that they warrant our trust – when they practitioners, processes, and products. In the scientific context, Heather identify and manage certain important epistemic risks. This unifies the Douglas (2004) has identified eight distinct meanings of process ‘ob- fragmented discourse on objectivity in the philosophy of science under jectivity’ alone, none of which is reducible to any other. And re- the heading of epistemic risk management, and thus resolves the am- searchers at the Essex Autonomy Project have identified at least six biguity that provides some of the fuel for deflationary responses. We distinct and mutually irreducible notions of objectivity in the law.1 endorse Koskinen's view and attempt to tailor it to the issues at stake in This ambiguity has provoked searing deflationary critiques in both the legal context. To do so, we develop a novel notion of phronetic risk, domains. To take a prominent example from the philosophy of science, and argue that we call practitioners, processes, and products of law Ian Hacking (2015) argues that objectivity is a useless “elevator con- objective when they identify and manage certain important epistemic cept” – by which he means a high-level philosophical concept, rather and/or phronetic risks. Our hope is that this will likewise bring unity to than a grounded one used by practitioners – and that we should stop an otherwise fragmented discourse, and, ultimately, help resolve the talking about it. To do worthwhile work on issues traditionally dis- kinds of disputes we mentioned at the outset. cussed under the heading of objectivity, he contends, philosophers should do ground-level analysis that targets the specific epistemic 2. Koskinen's epistemic risk account ‘vices’ that afflict researchers in their everyday practice, such asbiasor the corrupting influence of funding from interested third parties. The ideal of objectivity has an ontological and epistemological as- In the law, deflationary critiques have been even harsher. Catharine pect. The ontological aspect is concerned with the objective world, the MacKinnon (1991), for instance, argues that “objectivity” has long world as it is unmarred by subjective distortions—an “absolute con- served as a legal tool of oppression, because its lack of specific content ception” of reality (Williams, 1985) grasped, so to speak, from a God's allowed it to be co-opted as a stand-in for white male values. On her eye view, or “a view from nowhere” (Nagel, 1986). The epistemological view, objectivity not only lacks cognitive content, but it functions as an aspect, on the other hand, deals with the normative standards that instrument of injustice. govern our efforts to know the world. Despite these forceful deflationary critiques, the concept of objec- In keeping with a trend in recent philosophy of science to view tivity continues to do important work in scientific and legal practice. In science as a practice carried out by imperfect human agents for the sake science, we rely on a division of epistemic labour, and objectivity al- of human ends and interests, much recent work on the ideal of objec- lows that practice to persist. No one person can know everything, so we tivity sets the ontological aspect of objectivity aside. When you see divide the epistemic tasks between us (Kitcher, 2011, pp. 21–22; science as a practice organized around the interests of imperfect human Goldberg, 2011). To maintain this division of epistemic work, we need agents, the notion of “carving the mind-independent world at its mind- to trust that individual knowers and knowledge-production groups have independent joints” starts to look overly ambitious (Elgin, 2017, p. done their tasks well, and that we can trust their results. Objectivity is 151). Setting the ontological question aside, this recent work focuses the endorsement that we can trust the knowledge produced by others instead on constructing an epistemically normative ideal of objectivity (Grasswick, 2010; Schemann, 2011) and thus allows us to continue this via an analysis of the various ways actual scientists use the term “ob- vital social practice. Scheman (2011) and Grasswick (2010) speak of jectivity” in practice. We draw inspiration from Koskinen's (2018) im- larger audiences in the same terms: laypeople should be able to trust the portant work in this vein. outcomes of science. Objectivity assures them that such trust is war- Here are some of the ways Koskinen finds contemporary philoso- ranted. The epistemic authority of science thus rests on its claims to phers of science using the term objectivity: they treat objective claims objectivity (Koskinen, 2018; Reiss & Sprenger, 2017). as fallible; they think of objectivity not as an on/off concept but rather Objectivity serves similar functions in the law. The legal process as something that comes in degrees; they allow for great variety in what also depends on a division of labour wherein different legal actors can be objective, e.g., processes, people, claims, and communities; they perform distinct tasks. For that process to work, judges, juries, expert deploy the term differently across disciplines, projects, and contexts; witnesses and others must rely on each other to perform their tasks free and their efforts to ensure objectivity likewise vary according totheir from judgment-distorting factors like biases and conflicts of interests. current aims and context. To give a taste of this latter variety, Koskinen Objectivity is the endorsement that we can trust them to have done so. mentions three identified by Douglas (2004): 1) convergent objectivi- Moreover, objectivity also enables the wider public to place its trust in ty—when researchers reach the same results via different means, 2) the legal system. Hobbes (1651/1994) argued that the State exists so we procedural objectivity—when a process allows for one researcher to be can hand over our arduous individual claims to law enforcement to a replaced by another without altering the result, and 3) interactive ob- third party. We need not endorse the Hobbesian account to accept that jectivity—when a research community fosters lively and diverse critical we participate in formal legal processes, rather than taking them on exchanges. ourselves (like vigilante groups might), because we trust the legal Next, Koskinen argues that we can unify this diversity with a “risk system. In this vein, Postema (2001) argues that the law rests on le- account of objectivity.” Her argument begins with a basic fact of human gitimacy – on “the allegiance of all citizens” – and that this requires finitude: our imperfections as epistemic agents make our efforts toac- objectivity (p. 101). The normative practical authority of the law thus quire knowledge subject to risks of error, i.e., epistemic risks. In science rests on its claims to objectivity. there are many important epistemic risks, but the ones we have in mind In science and law, then, the concept of objectivity is ambiguous, when we talk about objectivity are risks of error that we are profoundly subject to deflationary critiques, and yet continues to do important and consistently prone to due to our frailties as epistemic agents, e.g., work. Given this shared predicament, we propose a shared rescue idiosyncrasies, illusions, cognitive biases, collective biases, and the like. strategy. Specifically, we extend Inkeri Koskinen's (2018) risk account Objectivity, according to Koskinen, assures listeners that they can rely on the products of science because “important epistemic risks arising from our imperfections as epistemic agents have been effectively 1 Members of the Essex Autonomy Project presented these findings in averted” (p.1). Thus, on Koskinen's account, when a speaker calls a February of 2017 at the Policy Institute at King's College for the first event scientific practitioner or a piece of science objective, she indicates that associated with the Mental Health and Justice project, a multi-disciplinary re- at least one of these important epistemic risks arising from our im- search initiative funded by the Wellcome Trust. Wayne Martin, Matt Burch, and perfections as epistemic agents has been recognised and measures have Sándor Gurbai prepared the public policy lab on objectivity. been taken to mitigate it. This doesn't mean that objective science is

61 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 infallible, just that best efforts have been made to avert certain factors “objective in the sense of being ‘mind-independent’ and causally effi- that are likely to take us further away from the truth. Sometimes, in cacious” (Leiter, 2001, p. 67). actual fact, scientists will still have gone wrong. However, it will still be Whether it's grounded in Dworkin's moral realism or Leiter's nat- your best bet to accept objective science, because it is more likely to be uralism, this approach to legal objectivity strikes us as overly ambi- right than the alternatives, where these kinds of epistemic risks have tious. If legal objectivity demands that we attend faithfully to the mind- not been managed. You should still rely on the science, even though it is independent moral and/or legal facts to determine uniquely correct fallible. answers to legal disputes, then it demands too much. Why think that The risk account of objectivity captures many of the ways that there's a uniquely correct determinate answer to nearly every legal historians (Daston & Galison, 2007) and philosophers of science question? The law, after all, is a contingent and frequently messy his- (Douglas, 2004) have identified scientists using the term. Objective torical practice organized around the evolving interests of imperfect claims are fallible, because our attempts to manage epistemic risks can human agents; and the situations to which we apply the law are often fail, e.g., we can overlook some epistemic risk, we can make an error in just as complicated and messy. Moreover, to understand objectivity as calculation, and so on. Objectivity comes in degrees because some determinacy is to treat it as an on/off concept—you either determine practice P1 can mitigate epistemic risk R better than practice P2, while the uniquely correct answer to the legal dispute, or you don't. And this both practices do a pretty good job. There's great variety in what can be leaves no room for the sort of comparisons, familiar to legal practice, objective, because processes, people, claims, communities, and many between more and less objective decisions, judges, and procedures. other things can avert important epistemic risks arising from our im- Moreover, it rules out the possibility that legal actors can make judg- perfect epistemic agency. The way we describe and achieve objectivity ments that are both objective and fallible. Like the view from nowhere varies significantly across disciplines, projects, and contexts, because in the philosophy of science, then, this looks like objectivity fit for gods, different situations pose different epistemic risks. Finally, our effortsto not limited imperfect agents engaged in a practice organized around ensure objectivity vary because different strategies avert different specific cultural and historical human interests. epistemic risks. For example, interactive objectivity – those communal Indeed, critics have reacted to the determinacy view of objectivity aspects of scientific processes, such as peer-review and scientific debate precisely along these lines. For instance, authors from Critical Legal – mitigate individual and collective sources of bias. Say an individual Studies like Roberto Unger (1986) have argued that if objectivity re- scientist has biased views on lung cancer that are fuelled by Big To- quires determinacy, then objectivity is beyond our reach, because bacco funding. The practices that constitute interactive objectivity will there's always room for rational indeterminacy when we seek to apply subject his claims to multiple sources of critical scrutiny and so make it an abstract body of law – in light of its institutional history – to the facts difficult for him get his biased views past the checks of the scientific of a particular case. Many scholars and legal practitioners agree: if community. According to Koskinen, what we see in the case of inter- objectivity in law depends on a) the existence of mind-independent active objectivity applies to every type of objectivity identified by au- moral and/or legal facts and b) our ability to know those facts, then it's thors like Douglas: each one manages particular epistemic risks. an unattainable ideal and we would be better off if we just stopped In this way, Koskinen unifies the fragmented discourse on objec- talking about it. tivity in science under the heading of epistemic risk management, and We propose another way forward. We agree with Leiter that we she thereby rescues the scientific conception of objectivity from pro- should take our cue from the natural sciences when developing a con- tracted disputes about its nature and merits. A similar rescue strategy is cept of legal objectivity (2001, p. 67); but our agreement stops there, appropriate for objectivity in the law, but it requires that we think more because we think Leiter latches on to the wrong model of scientific expansively about the risks at play. objectivity. Instead of pursuing an analogue to the view from nowhere, we should follow the lead of more recent work in philosophy of science. 3. A risk account of objectivity fit for law We should set aside ontological questions about some objective realm of moral and/or legal facts and concentrate instead on developing an Much of the philosophical discourse on legal objectivity treats the epistemically and practically normative account of objectivity. And we law's objectivity as a question of its determinacy, maintaining that the should develop that account not in terms of some abstract conception of law is objective insofar as it determines “uniquely correct outcomes” in the law but rather in relation to the ways actual legal practitioners use actual or hypothetical cases (Brink, 2001, p. 65). Put otherwise, legal the term objectivity. To develop such an account, we adapt Koskinen's decisions are objective if and only if they achieve the result that the law risk account of objectivity to the legal context. really requires. This notion of ‘the one determinate correct answer’ Given her interest in scientific objectivity, Koskinen naturally fo- serves as legal philosophy's analogue to the view from nowhere in cuses strictly on the epistemic risks that objective science averts. But philosophy of science. The best-known version of this approach is Ro- objectivity isn't a purely epistemic ideal. It also plays a role in a range of nald Dworkin's “right answer thesis” which maintains that most legal other contexts – e.g., morality, politics, and law (Nussbaum, 2001; cases – even hard cases – have objectively correct determinate answers. Shafer, 2003; Sibley, 2001) – where the stakes are decidedly non- And, as Dworkin (1986) argues in Law's Empire and elsewhere, the epistemic. So when we talk about legal objectivity, we mean something correct answer to a legal dispute offers the best fit with the law's in- broader than epistemic risk management. stitutional history and the best moral justification of that history. Thus, How should we think of the non-epistemic risks posed by the legal for Dworkin, objectivity in law depends not just on the objectivity of context? Biddle and Kukla (2017) provide a clue. Although they also legal interpretation but also on the objectivity of morality, i.e., “an focus exclusively on knowledge production, Biddle and Kukla identify a independent, subsisting realm of moral facts” (Dworkin, 1996, p. 97). broad range of risks that are not strictly epistemic but nevertheless bear Other legal philosophers defend the view of legal objectivity as on our epistemic practices. Philosophers of science have long-accepted determinacy, while rejecting the link Dworkin forges between law and that moving from a body of evidence to a hypothesis involves risk – the morality. Brian Leiter, for example, argues that if we tie objectivity to so-called problem of ‘inductive risk’ (Rudner, 1953). Biddle and Kukla moral realism, few will buy it, because an absolute conception of argue further that risk in knowledge production is more ubiquitous, morality is a hard sell. Leiter still insists that an objective conception of involving everything from model selection through to how particular law requires determinate answers, but he proposes a naturalistic al- phenomena are classified. They describe this more expansive notion of ternative to Dworkin's approach. According to Leiter, we should base risk as ‘phronetic risk’, which they define as “epistemic risks that arise our determinate correct answers to legal questions not on non-natural during the course of activities that are preconditions for or parts of moral facts but rather on a “metaphysically objective” realm of “legal empirical (inductive or abductive) reasoning” (p. 220). Such risks arise, facts” (Leiter, 2002, p. 969) that, like the objects of nature, are then, when researchers must deliberate in light of their values and

62 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 interests in order to make calls that are fateful for the balance of the available reasons best support; and d) discover and know how to epistemic risks. secure efficient means to achieve that course of action (or setofac- We want to expand this category of phronetic risk to tailor tions). And she must do all this in a way that is sensitive to and con- Koskinen's risk account of objectivity to the legal context. When we strained by the relevant moral considerations. This is the work of speak of objectivity in the law, we are most often concerned not just phronēsis. It is the capacity to make context-sensitive calls about the with accurate knowledge acquisition strategies but rather with the best course of action across a range of variable, dynamic concrete cir- normative requirements of practical reason. Consider a few examples. cumstances.3 Such calls are risky, because even the most skilful prac- We claim that divorce proceedings are objective when the legal deci- tical reasoners can get them wrong. Moreover, going wrong in such sions that determine the outcome treat all relevant parties fairly; ob- cases is not a matter of failing to formulate accurate knowledge claims; jectivity here thus averts the risk of our practical reasoning failing to rather, it's a matter of failing to discern what the situation calls for track the normative requirement of fairness that we associate with morally and practically speaking. Thus, the risks that threaten our justice. To take another example, we say that a judge's decision fails to ability to get this sort of call right are not epistemic but rather phro- be objective when it's shaped by racial bias; objectivity in his case netic. would have averted the risks of personal bias and racial animus af- If epistemic risk is “any risk of epistemic error that arises anywhere fecting his reasoning and decision-making. Finally, when a witness on during knowledge practices” (Biddle & Kukla, 2017, p. 218); phronetic Big Pharma's payroll testifies to the moral integrity of the CEO who runs risk, as we conceive it, is any risk of error that arises during practical her company, we might doubt her objectivity, as objectivity on her part reasoning about the correct judgment to make and the best course of would require her to prevent a considerable conflict of interests from action to take. Just as epistemic risks represent a hurdle to science distorting her appraisal of her boss's character. In these examples, we achieving its theoretical aims, especially its paramount aim of truth, see legal actors running the risk of getting things wrong, but the risks at phronetic risks pose a hindrance to the law achieving its guiding play, at least principally, are not epistemic. For the most part, we're not practical aims, especially its highest aim of justice. Moreover, just as worried that an attempt to produce knowledge will go wrong in these objective science continually faces and mitigates diverse epistemic cases; rather, the relevant risks represent threats to practical reasoning. risks, objective legal practice continually faces and mitigates diverse These examples are undoubtedly concerned with phronetic risk, then, epistemic and phronetic risks.4 Finally, in the scientific context, we saw but not of the sort of epistemic-phronetic risk that Biddle, Kukla and that objectivity is concerned with managing important epistemic risks Koskinen have in mind. These examples from the law are not targeting a that we're subject to profoundly and consistently because they arise sub-category of epistemic risks, but rather a class of risks that are from our imperfections as knowers. The same is true in the case of strictly phronetic, i.e., risks of getting things wrong in our practical phronetic risks: the ones objective legal practice strives to avert arise reasoning. from our imperfections as practical reasoners. We thus need to extend the notion of phronetic risk beyond epis- As knowers and deliberators, we aim to get things right in theory temic concerns to cover the wider range of risks encountered in the and practice. To do so, we must avert the important epistemic and practical reasoning of the legal system. To articulate this broader notion phronetic risks that stand in our way. Some of these important epis- of phronetic risk, we turn briefly to Aristotle's Nichomachean Ethics. In temic and phronetic risks arise from our imperfections as epistemic and that work, Aristotle identifies five intellectual virtues. One of themis practical agents. Objectivity is the assurance that at least some of these epistēme, which, as most readers will know, is typically translated as latter risks have been managed.5 knowledge (or scientific knowledge), and from which the term epis- With this expanded account of phronetic risk in view, we propose temic derives. Hence, epistemic risks threaten our efforts to acquire that objectivity in the law is the assurance that measures have been knowledge. Another intellectual virtue Aristotle names there is taken to manage the epistemic and phronetic risks of error inherent in phronēsis, which is typically translated as practical wisdom or practical legal practice. Objectivity is a self-responsible stance that strives to intelligence, and from which we derive the term phronetic. Phronēsis identify and avert epistemic and phronetic risks that arise from our does a lot of important work in Aristotle's virtue ethical theory, but we imperfections as agents. Such a risk account of legal objectivity will will narrowly tailor our discussion of the concept to our current con- have all the virtues of its scientific corollary. First, it allows for objec- cerns. tive legal claims to be fallible, which is attractive given the inherent Phronēsis denotes the human capacity for “concretely situation- fallibility of human judges and other legal actors. Second, it entails that specific discernment” (McDowell, 2007, p. 340). According to Aristotle, objectivity is not an on/off concept but rather comes in degrees, which without this capacity one cannot live a virtuous life. For, he argues, allows us to compare judges, procedures, and outcomes along a con- even if we get lucky, and we're naturally endowed with virtuous im- tinuum of objectivity. Third, it allows us to account for the great variety pulses, and our upbringing habituates us to seek the good and teaches of things in the law that are said to be objective: evidence; expert us our culture's code of conduct, without phronēsis, we will still be apt witnesses; judges and their decisions; the jury and its verdicts; proce- to go wrong with respect to the demands of virtue. Why? Because doing dures; standards; legislative bodies; and the legal system itself. These the right thing isn't simply a matter of rule-following; as Aristotle things can all be objective, but they are so in their own way, because (2000) famously argues, right action requires that we act “at the right times, with reference to the right objects, and to the right people” (110b21–22). Think of it this way. You can teach a child with a good 3 For a lucid discussion of the intellectual virtue of phronēsis, see McMullin disposition and good habits that justice is “the constant and perpetual (2018, pp. 110–114). will to render each his due” (Institutes of Justinian)2; but you cannot 4 Indeed, in most cases the law must avert many phronetic risks as once. For thereby expect the child to know how to render each person her due instance, in the child protection cases that Munro and Hardie (2018) in their across the vast range of diverse and evolving concrete circumstances recent work on objectivity, when the Family Court makes assessments about that life presents us with. Such situation-specific discernment requires whether an at-risk child should remain with their family or be placed in care, the Court must simultaneously manage risks that have to do with our limited the practical intelligence to a) move from general rules to specific cir- abilities to forecast future outcomes, the social value we place on keeping fa- cumstances; b) descry the morally salient features of novel and ever- milies intact, the potential physical harms to the child, and to treating all the shifting situations; c) discern which course of action (or set of actions) relevant parties (parents, the child, other caregivers) with appropriate concern, amongst other risks. 5 From here on, when we refer to important epistemic/phronetic risks, 2 Cited in Miller (2017). The Institutes of Justinian is a codification of Roman readers can assume that we mean important epistemic/phronetic risks that arise Law from 6th c. AD. from our imperfections as epistemic agents/practical reasoners.

63 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 they avert different epistemic and/or phronetic risks. Finally, this ap- legal norms, moral principles, and the law's institutional history to bear proach allows us to tailor our understanding of what objectivity de- on the concrete particulars of a case. Like the move from data to in- mands to our current aims and context, because the epistemic and/or terpretation in science, this hermeneutic move from the law to the facts phronetic risks we face will vary with our aims and context.6 Im- of the case is shot through with uncertainty. Not only is there often a portantly, the language of ‘phronetic risk management’ is not intended low probability of a unique, determinate, and correct interpretation of a to imply anything approximating a ‘check boxing’ approach to risk, or case in light of our legal norms, moral principles, and the actual in- any managerial bureaucratic oversight. Rather, drawing on our earlier stitutional history of the law. But it is even less likely that an imperfect discussion of Aristotle, we conceive of this as a type of practical limited agent will actually know what that uniquely correct inter- wisdom. pretation is. In the words of former Chief Justice of England, Lord In the second half of this paper, we change pace somewhat. We Bingham, “when you are deciding a case you usually feel that there is a follow Koskinen's procedure by identifying some of the products, per- choice of answers… …To say that there is one right answer and one sons, and processes that actual legal practitioners deem objective and wrong answer is just not at all how it feels.”8 This uncertainty makes indicating which epistemic and/or phronetic risks they avert. There are the move from the law to the particulars of a case risky. The move undoubtedly more types of objectivity and associated risks in the law almost always involves ethical risks – i.e., risks of harm to others – than we will identify here, and still more will likely emerge in the fu- because legal decisions tend to have real-world consequences. More- ture. After all, as Koskinen notes, the risks a community deems salient over, moving from the body of law to a particular case also always change over time. Our aim here is not to offer an exhaustive list, but involves phronetic risks: we can go wrong with respect to the normative rather an indicative one that illustrates how a risk account of objectivity practical requirements of the law. For example, there's the phronetic would function in the legal context. We did not, however, construct our risk of interpreting the law in a way that leads to an unjust decision, or list at random; rather, we selected certain types of objectivity and a decision that is less just than a viable alternative would be. epistemic/phronetic risks due to their prevalence in current legal Objective judges acknowledge these phronetic risks inherent in the practice. Indeed, we think it's hard to imagine how a legal system could hermeneutic process and take steps to avert them. This includes relying make claims to objectivity without managing the epistemic and phro- on techniques that are analogous to the interpretative norms of science netic risks we discuss below. To continue our comparison between just discussed (Eskridge & Ferejohn, 1994; Knight & Johnson, 1994). scientific and legal objectivity, we will introduce each type oflegal One example of an interpretative strategy designed to mitigate phro- objectivity with a quick look at its scientific analogue. netic risk is the principle of stare decisis, or the requirement that judicial precedent be given interpretative priority. When judges prioritise pre- 4. Types of objectivity and the risks they manage cedence in this way, we have more reason to trust them to interpret the law in similar ways in similar cases (Foster, 2008). This averts an ob- 4.1. Hermeneutic objectivity vious ethical risk—for it would be unjust to hold people accountable to unpredictable laws. But the practice also manages phronetic risks, e.g., The contemporary philosophical concern with epistemic risk man- when judges are tempted to make idiosyncratic calls or to “legislate agement began with the problem of inductive risk, i.e., the problem of from the bench” on the basis of pet ideologies, the practice of prior- moving from a body of data to a unique interpretation (Rudner, 1953). itising precedence anchors their reasoning in the common law and a That move involves potential ethical, phronetic, and epistemic risks. shared history of interpretation that tempers their personal take on For example, imagine a public health context where accepting hy- things. Again, this is an instance of phronetic, rather than epistemic, pothesis H1 might lead to 1000 deaths, while accepting H2 might lead to risk management—we don't expect that adhering to precedent will get 1000 people experiencing slight discomfort. Accepting either hypoth- us closer to the ‘true’ interpretation of the law, only that it guards esis involves ethical risks, i.e., risks of harm, but we would demand a against risks like idiosyncrasy and individual bias. higher degree of certainty to accept H1 because the ethical stakes are higher. Moreover, the move from data to interpretation in this case also 4.2. Procedural objectivity involves managing phronetic risks, because we have to make a practical moral judgment about the balance of those ethical risks, and we could Earlier we touched on procedural objectivity in science, i.e., when a get that call wrong. Finally, there are also epistemic risks to consider. process allows for one researcher to be replaced by another without The very fact that we are not dealing with deduction but rather making altering the result. With procedural objectivity, scientists rely on an inductive move that involves degrees of probability makes us vul- methodological procedures to avert the epistemic risks of idiosyncrasy nerable to the risk of epistemic error. One of the ways objective science and individual bias. Daston and Galison (2007) famously describe manages such epistemic risks is via interpretative norms. For instance, ‘mechanical objectivity’ along these lines: consider the standardisation of p-values, i.e., the standard that 5% significance is required to reject the null hypothesis and accept your By mechanical objectivity we mean the insistent drive to repress the own hypothesis. As Parascandola (2010) argues, such p-value stan- willful intervention of the artist-author, and to put in its stead a set dardisation is an epistemic risk management strategy.7 The scientific of procedures that would, as it were, move nature to the page community agrees on a standard p-value to offset worries about un- through a strict protocol, if not automatically certainty in each new case. (Daston & Galison, 2007, p. 121). An analogue to inductive risk runs through the practice of legal The idea, then, is that such procedures eliminate distortions that interpretation. When judges decide a case, for instance, they must bring stem from the scientist's own subjective contribution to her observa- tions and thereby enhances her focus on the objects she observes. 6 “Context” in the law means that what will be relevant will vary on a case-by- Moreover, the fact that any other scientist could follow the procedure case basis. Philosophy of science typically draws a distinction between the and attain the same results suggests that it successfully manages the context in which knowledge is produced and the context in which it us used epistemic risks associated with subjective distortions like idiosyncrasy (Montuschi, 2016, 60), but given the deeply applied nature of the law, all and individual bias. If multiple researchers reach the same result, then contexts are contexts of use. it's less likely that any idiosyncrasies or biases specific to them as 7 Parascandola goes on to argue that p-value standardisation creates addi- tional risks, and so is ultimately not successful. But our point was to highlight how scientists use interpretative norms to manage epistemic risks, not to assess 8 This remark is cited in Guest (2007, p. 100) and comes from Susskind the value of a particular standard. (2005).

64 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 individuals have distorted that result. Finally, strict protocols also manage the epistemic and/or phronetic risks that individual practitioners prevent the scientist from inadvertently introducing a change to the might intentionally or inadvertently contaminate evidence or impose their process that affects the outcome. personal quirks and biases on the process. In the law, procedural objectivity plays an analogous role. Some legal procedures principally aim to manage epistemic risks. The Law of 4.3. Objectivity as independence of judgment Evidence, for instance, focuses on averting epistemic risks by ensuring that legal actors follow appropriate procedures in acquiring the evi- Another strategy for managing epistemic risk in science is the attempt dence that appears before the Court. As Alvin Goldman (2005) notes, to maintain independence of judgment. It's well known that funding from it's generally held that one of the principal aims – if not the principal interested third parties tends to skew research results towards those par- aim – of evidence-handling procedures is to promote “the accurate or ties' interests; and, correlatively, we also know that the best evidence tends truthful determination of facts relevant to the case at hand” (p.164). For to come from studies conducted by scientists who operate independently example, proper procedures of evidence management require that re- of such interested parties. For example, as Stegenga (2018) notes, the most levant parties handle evidence in a way that leaves no significant room accurate and reliable randomized controlled trials and systematic reviews for doubt that someone could have accidentally altered or deliberately to test medical interventions are typically carried out by “academics who tampered with the evidence. The analogy to procedural objectivity in are independent of the manufactures of the medical interventions in science is clear. Just as a strict experimental protocol is designed to question” (p. 2). When researchers have an interest in the outcome of prevent an individual scientist from introducing changes to the process scientific research, they tend to err on the side of that interest, which quite that affect the results, following a strict evidence-handling procedure in often takes them away from the truth. The scientific community thus criminal cases assures us, to a reasonable degree, that no one has cor- promotes strategies designed to avert this epistemic risk, such as requiring rupted or substituted the evidence in a manner that leads us away from researchers to declare their funding sources and other conflicts of interests. a truthful determination of the facts. In our terms, then, these proce- These strategies aim to mitigate idiosyncrasy, bias, and motivated rea- dures contribute to the law's objectivity by helping us avert important soning, and to encourage other members of the scientific community to epistemic risks.9 closely scrutinise the research results. Other parts of the law use procedures predominantly to manage The law also treats independence of judgment as an essential strategy phronetic risks. The proceduralisation of the judicial process aims to to manage phronetic risk. Notice what Waldron puts second on his list of allow for one legal actor to be replaced by another without significantly indispensable procedural characteristics: a “legally-trained judicial officer, altering the result. This provides a useful lens to understand procedural whose independence of other agencies of government is assured”. This rights and our highly proceduralised court hearings. For example, aims to assure the person before the law that the outcome will not be consider some of the procedural characteristics that Jeremy Waldron predetermined by third party interests that are normatively irrelevant to (2011) deems indispensable to the rule of law. According to Waldron, the adjudication of her specific case. But independence of judgment does the government should not impose any “penalty, stigma or serious loss” not just require distance from the interests of third parties; it also requires on someone without procedures that involve: that persons involved take measures to prevent their own personal inter- ests from affecting their judgment. We can understand anti-sympathy in- A. A hearing by an impartial tribunal that is required to act on the basis struction for jurors in this light. For example, consider California v. Brown of evidence and argument presented formally before it in relation to (1987) wherein the trial court instructs jurors not make their decision legal norms that govern the imposition of penalty, stigma, loss etc.; based on “mere sentiment, conjecture, sympathy, passion, prejudice, B. A legally-trained judicial officer, whose independence of other public opinion or public feeling.” We see a similar concern addressed by agencies of government is assured; the Court of Protection of England and Wales in CC v KK and STCC (2012), C. A right to representation by counsel and to the time and opportunity as Justice Baker writes, required to prepare a case; D. A right to be present at all critical stages of the proceeding (2011, p. [T]here is a risk that all professionals involved with treating and 6).10 helping that person – including, of course, a judge in the Court of Protection – may feel drawn towards an outcome that is more pro- These procedural characteristics clearly aim to manage the phronetic tective of the adult and thus, in certain circumstances, fail to carry risks of making biased, discriminatory, and/or arbitrary decisions. The out an assessment of capacity that is detached and objective. On the basic idea is that no matter who you are, and whatever your personal other hand, the court must be equally careful not to be influenced by characteristics, the law's procedures exist to ensure that you receive the sympathy for a person's wholly understandable wish to return home same treatment as anyone else not relevantly different from you. In other (par. 25, emphasis added). words, such procedures are designed to prevent legal actors from imposing This passage identifies two potential personal interests – the urgeto their idiosyncrasies and biases on the legal proceedings. In both the sci- protect a person and sympathy for her wishes – that, in our terms, pose entific and the legal context, then, the emphasis on procedures isawayto phronetic risks that judges and others involved in such cases must manage with what Justice Baker characterizes as a detached and ob-

9 jective stance. Again, independent judgment requires independence not Of course, as Goldman also points out, not all evidence-handling procedures just from the interests of third parties but also from one's own interests, aim at truth. For instance, exclusionary rules make illegally obtained evidence however noble or sensitive their underlying motives may be. inadmissible, even though such evidence might lead to a more truthful de- Whether motivated by group-interest or self-interest, conflicts of termination of the facts. This does not, however, make the law less objective. It just highlights the fact that strategies designed to manage epistemic risks must interest are inimical to independence of judgment, as they motivate the simultaneously manage relevant ethical and phronetic risks. Courts do not agent to decide on behalf of her group or herself, rather than attending admit illegally obtained evidence because of the associated ethical risks, e.g., to normatively relevant facts, reasons, and arguments. If a judge has a doing so might incentivize criminal behavior amongst people trying to win a personal or social stake in the outcome of a case, those involved might case. However, in defining what constitutes illegally obtained evidence, thelaw naturally worry about her potential lack of impartiality.11 If a juror also has to manage phronetic risks, e.g., the risk of making the wrong call when can't regard the defendant without prejudice, then the latter will not be balancing the value of a more truthful determination of the facts against the ethical risks just mentioned. 10 This is only part of Waldron's list but we thought it sufficient to make our 11 For discussions of the problem of conflicts of interest in relation to objec- point. tivity see Harding (1992) or Resnik (1998).

65 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 able to trust the former's reasoning and ability to follow the evidence need to take a cue from Feminist Philosophy of Science and its emphasis where it leads. Procedures, protocols, and instructions designed to on perspectival diversity. To achieve a truly robust deliberative con- promote independence of judgment thus contribute to the law's objec- sensus, it's not enough that a claim be “maximally supported by the tivity by managing the phronetic risks of partiality and conflicts of in- arguments and the balance of reasons available for articulation and terest that could lead the process of legal reasoning away from a nor- assessment by reasonable and competent persons in a fully public de- matively correct judgment. liberative process” (Postema, 2001, p. 117). Such maximal argu- mentative support is not enough because which reasons are “available 4.4. Objectivity as publicity, or deliberative objectivity for articulation and assessment” hinges crucially on which people we include in the deliberative process. Again, critical exchange might Earlier we mentioned interactive objectivity in science, which ob- eliminate idiosyncrasy and individual bias, but it's unlikely to root out tains when a research community fosters lively and diverse critical collective bias unless the group hears the voices of outsiders. Otherwise, exchanges. This type of critical, argumentative activity aims to mitigate the homogeneity of legal practitioners can obscure collective biases at risks of idiosyncrasy as well as individual and collective biases work in the legal system. (Longino, 1990). It's clear enough how critical dialogue manages idio- Consider, for example, the case of Buck v. Bell (1927) in which the syncrasy and individual bias: it allows the community to identify and United States Supreme Court upheld a Virginia circuit court ruling that eliminate merely personal takes on the available evidence. How a cri- the state should sterilise Carrie Buck, a young, poor, single mother, tical exchange uncovers collective bias, however, is less obvious. Col- falsely alleged to be “feebleminded.” At the time, the Supreme Court lective biases, after all, are shared, and so they tend to operate as had several enthusiastic supporters of eugenics on the bench, including background assumptions that shape the group's conversations; however Oliver Wendell Holmes, who wrote the Court's opinion. Under such critical their exchanges may be, then, it's unclear what mechanism conditions, a critical evaluation of available reasons and arguments falls could reliably bring such shared background assumptions into view. short. Had the conversation included the voices of women, people This highlights the importance of outsiders for interactive objec- living in poverty, and/or people with disabilities, the Court's collective tivity, a point that has gained theoretical prominence with the devel- bias may have been exposed. Maximal argumentative support is not opment of Standpoint Theory in Feminist Philosophy of Science. enough; diversity of perspectives is also essential to making objectivity Standpoint theories, such as Harding (1991) and Wylie (2003), hold as publicity properly public. Of course, such diversity will never guar- that the “view from nowhere” or the so-called “value-free ideal” is antee the elimination of collective bias—such a conceit would only lull impossible. Instead, they argue that all knowledge is from a particular us back into the dreamy stupor of the illusory view from nowhere. But perspective and advocate for more diverse scientific communities, be- the voices of outsiders remain our best hope for catching sight of the cause looking at the evidence from different social and political per- blind spots that lie hidden in consensus views. This captures some of the spectives may allow for different factors to become more or less salient impetus behind the well-known slogan and principle of disability rights (Wylie, 2003). As an example of this, consider the case of early AIDS activism, “Nothing About Us Without Us”; many activists believe, science as described by Steven Epstein in Impure Science (1996). AIDS rightly we think, that conversations about disability rights will be in- was initially aetiologically mysterious, but the initial framing of AIDS flected with collective biases that persons with disabilities are uniquely was that it must be caused by some aspect of the ‘gay-lifestyle’ (rather positioned to identify, challenge, and correct.12 than being an infectious disease with a microbial cause), focusing in on Despite the analogy between interactive and deliberative objec- the most sensationalised aspects of the lifestyles of men in the ‘urban tivity, we should highlight the fact that the former principally manages American gay-scene’ (e.g., promiscuity and drug use), and thus dis- epistemic risks, while the latter mitigates phronetic risks. That is, the missing evidence that ran counter to the hypothesis that it must be life- former helps us avoid error in knowledge production, while the latter, style related; such as the monogamous gay men, heterosexual men and as the Buck case illustrates, manages the risk of our practical judgments women, and children who presented with AIDS (Epstein, 1996, pp. falling short of the requirements of justice.13 48–50). Ultimately it was microbiology and the discovery of the HIV The intuition behind deliberative objectivity is clear. It rests par- virus, not diversifying the scientific community, that put us on the right tially on the notion that one's judgments improve when they must be causal track, but this provides a cautionary tale of the dangers of defended against critical scrutiny, and partially on the intuition that looking at the evidence from one social/political perspective. Thus, the one of the virtues of objectivity is transparency. Even Munro and Hardie example constitutes a negative illustration of the importance of inter- (2018), in their criticism of objectivity, accept that the transparency active objectivity. Consider a related positive illustration: later in the associated with the term is worth striving for, because it allows others 1980s, when treatment first became available, mainstream science had to check that no biases, idiosyncrasies or malevolent motives have to consult members of the AIDS activist community, because they had warped the reasoning process. become the experts, as those living with the disease, and this per- spectival knowledge was essential for successfully developing treatment (Epstein, 1996). 4.5. Structural objectivity In the legal context, the analogue to interactive objectivity is what legal philosophers call ‘objectivity as publicity’ or ‘deliberative objec- Like interactive and deliberative objectivity, structural objectivity tivity.’ On Postema's (2001) compelling account of this form of objec- also manages a group-level phronetic risk, but not one that diverse tivity, a judgment is objectively correct if and only if it can be justified critical exchanges can do much to manage, i.e., ‘structural biases’. by public practical reason. The basic idea is that deliberative objectivity Structural epistemic biases refer to the ways in which the very structure holds for practical reasoning procedures and their outcomes when they of the scientific research environment leads us to neglect the perspec- are acceptable to all (in practice or in principle). Thus, we achieve tives, stories, and interests of large swathes of society and thereby deliberative objectivity when we arrive at judgments through a public, skews our knowledge base. To be clear, this is not a problem that is deliberative process that fully considers and assesses the available and necessarily solved by just making the research community more normatively relevant arguments and reasons. Under the best circum- stances, we can see this kind of deliberative objectivity at work in the 12 See Charlton, 1998. jury process, law consultations, and legal argumentation. 13 Miranda Fricker's Epistemic Injustice (2007) shows that exclusions of the Like interactive objectivity, deliberative objectivity aims to mitigate views of members from marginalised groups might not just be an epistemic loss, risks of idiosyncrasy as well as individual and collective biases. To do so but may also be an injustice to those excluded. And so even focussing just on the effectively, however, accounts of deliberative objectivity in thelaw epistemic may sometimes also involve considerations about justice.

66 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 diverse: diverse scientific research communities can (and many do) People who fight for structural objectivity in the law attempt to change work in structurally biased research environments. Consider, for ex- how the law works, so that it doesn't systematically favour privileged ample, the case of neglected diseases, as discussed by Reiss and Kitcher groups of people. On this view, then, the law exhibits structural objectivity (2009). Neglected diseases are those that receive a proportionally to the extent that it treats everyone the same unless they are relevantly smaller share of the available research and biomedical resources re- different.14 Why is this a matter of objectivity at all and not strictly an lative to their contribution to the global disease burden. Note: ethical issue of non-discrimination? The basic idea is that a discriminatory legal system is shaped in light of – and so biased towards – a privileged For instance, malaria, pneumonia, diarrhea and tuberculosis to- form of subjectivity. Moreover, just as a structurally biased research en- gether account for 21% of the global disease burden, but receive vironment will skew our body of knowledge towards the interests of pri- only 0.31% of all public and private funds devoted to health re- vileged groups, so will a structurally biased legal system skew our body of search legal judgments towards the advantage of some groups and the dis- (Reiss & Kitcher, 2009, p. 264). advantage of others. Whereas a legal system wherein “all persons are equal The real-world consequence of this is that many thousands of people before and under the law and are entitled without any discrimination to die from diseases that almost exclusively afflict the poorest parts ofthe the equal protection and equal benefit of the law” (UN General Assembly, world, many of which have been completely eradicated from more af- 2006, Article 5.1.) is objective in the sense that it is not biased towards the fluent regions. Tuberculosis, for instance, results in 1,566,000 deaths interests of a particular form of subjectivity, and so it should not produce a annually, despite being all-but-eliminated from richer portions of the body of judgments that systematically helps certain groups and hinders world (although, it is still a problem amongst socially and economically others. An objective legal system roots out structural biases and treats marginalised segments of wealthy countries, such as the United everyone the same regardless of their personal characteristics, and by Kingdom). One major reason that biomedical research neglects diseases doing so, it manages some of the deepest risks that legal reasoning will run that mostly affect the poor is that large pharmaceutical companies afoul of the demands of justice. undertake the majority of that research. The cost of taking a new drug from initial research to shelf is astronomically large (around $800 4.6. Objective evidence million for a single new drug). Since pharmaceutical companies could never expect to make back this investment in the world's poorest re- Finally, we turn to “objective evidence”, a common expression in sci- gions, they do not undertake research that is relevant to these places entific and legal discourse. Although the nature of evidence isamajor (Reiss & Kitcher, 2009, pp. 265–267). This is a structural problem that philosophical issue in its own right, we only wish to sketch the link be- is unlikely to be addressed by the influx of researchers from more di- tween objective evidence and epistemic/phronetic risk management. verse backgrounds. First, note the basic ambiguity in the scientific use of the expression In addition to being an obvious ethical concern, this poses an “objective evidence”. Sometimes we use the expression as if it refers to a epistemic risk, in that we end up with medical knowledge skewed to- particular kind of evidence; as though the evidence were objective by its wards one very small segment of the world's population. Moreover, very nature. What we typically mean by this is that the evidence is publicly interactive objectivity can't manage this risk—introduce scientists from available.15 Such evidence is objective, in our sense, because its public diverse backgrounds into this field of medical research and there still availability manages certain epistemic risks. Because the evidence is out won't be money for research on neglected diseases. The main safeguard there for all to see, we can come at it from different perspectives and against this kind of epistemic risk is structural objectivity—the attempt appeal to it as a neutral arbiter when adjudicating disputes. And this in- to change the research environment so that it's not systematically tersubjective triangulation on a piece of public evidence mitigates the skewed towards the perspectives, stories, and interests of privileged epistemic risks of idiosyncrasy and personal blind spots. To be clear, we members of society. Reiss and Kitcher (2009) suggest that the way to do are not just talking about procedural or interactive objectivity again. It's it in the case of neglected diseases is to implement Kitcher's suggestions not the researchers' activity alone that makes the evidence objective. for ‘well-ordered science’, as outlined in his book Science, Truth and There's something about the evidence itself that makes it objective, Democracy (Kitcher, 2001). The details of this suggestion are too namely, its publicity. Such evidence is objective in a way that, say, purely lengthy for us to spell out here, and not directly relevant to our ends. private mental experiences are not. Multiple people can see public evi- The point is that philosophers of science take seriously the risk of dence from different perspectives, and so the very nature of such evidence structural bias, like that seen in the case of neglected diseases, and they mitigates the risk of certain kinds of error. have suggestions for how we might manage that risk. But we also speak of objective evidence in at least one other way: we In the law, we see an analogous strategy at work, with feminist phi- refer to the results of scientific inquiry – the products of science –as losophers once again leading the way. Prominent feminist critiques argue objective evidence. For example, when asked for objective evidence for that the law is not objective because it exhibits structural biases, system- the claim that X causes Y, we will likely appeal to empirical research atically subordinating the interests of women to those of men (Mackinnon, results meant to demonstrate that causal link, e.g., “You want objective 1983, 1987). Mackinnon, for instance, argues that the ideal of objectivity evidence? Here are 30 RCTs linking X to Y.” In this case, as Elgin (2017) is in fact the male point of view masquerading as a “nonsituated, universal puts it, “What justifies calling a particular result objective is that itis standpoint” (1983, p. 636). The basic claim behind her critique is that the the product of an objective procedure” (p. 159). In our terms, the re- very structure of the law privileges powerful white men and disadvantages sults of scientific inquiry can count as objective evidence in favourof women, ethnic minorities, persons living in poverty, and persons with some claim, because those results were reached by processes that avert disabilities; and the law deploys the rhetoric of objectivity to cover over this systemic oppression. If this legal order is simply what objective, value- 14 neutral reason demands, then it cannot be unjust. In this way, Mackinnon Relevantly different here means that a person is different in such awaythat argues that the language of objectivity “reinforce(s) existing distributions treating her like everyone else would constitute unfair treatment, e.g., failing to provide a person with intellectual difficulties with reasonable accommodations of power” (1983, p.645). Authors working in Critical Legal Studies (Unger, in her education would constitute a failure to treat differently someone who is 1986), Critical Race Theory (Crenshaw, Gotanda, and Peller, 1995), and relevantly different. Disability Studies (Wolbring, 2008) have similarly argued that the law is 15 For Conee and Feldman (2008) scientific evidence is “publicly available” not just structurally sexist but also racist and ableist. Some think this by definition (p. 84). We are not sure that all scientific evidence mustbe critique recommends the view that the very ideal of objectivity is corrupt; publicly available—after all, introspective reports are private, yet they still but others see it as a call to a deeper and more adequate notion of ob- serve as evidence in psychological science. But we agree that publicly available jectivity. evidence is especially important in science.

67 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 important epistemic risks. What sort of processes? Exactly the sort sufficient. We acknowledge the validity of her claim that objectivity can we've discussed thus far: lively critical exchanges, peer review, trian- function as a tool of oppression, but we don't think it necessarily must. gulation, practices that root out conflicts of interest, critical scrutiny of Moreover, we see authors like MacKinnon as allies in the process of collective and structural biases, strict experimental protocols, and so managing epistemic and phronetic risks. Such critics highlight collec- on. Again, we're not just repeating what we've said about other forms of tive and structural biases at work in the law that we must manage if we objectivity here; rather, we're pointing out that results reached via these hope to have an objective and just legal system. Far from deflating various strategies of epistemic risk management earn the label of ob- objectivity, then, such critics do important work to make it possible. jective. Objective evidence in this second sense denotes results we can What about Hacking's deflationary worries? Does the term ‘objec- rely on because they were reached via strategies that manage epistemic tivity’ really do any work on our approach? In other words, does talk risk; and we should rely on those results to a degree that matches the about objectivity do anything over and above simply pointing to the rigor and known reliability of those risk management strategies. Thus, totality of risks discussed in section four? Why talk about objective we have i) evidence that is objective in virtue of its public nature and ii) evidence, for example, rather than simply saying that in selecting and evidence that is objective in virtue of the epistemic risk management gathering our evidence we've managed the known epistemic risks? Why strategy by which it was reached. talk about deliberative objectivity rather than saying that normatively Objective evidence in the law reflects a similar ambiguity. Much like correct deliberative procedures were followed? Why not just stop the scientific context, the preferred and most powerful form of evidence talking about objectivity and focus instead on doing ground-level ana- in legal practice is publicly available evidence. For example, in criminal lysis of the specific epistemic virtues and vices in individual instances of proceedings, prosecutors tend to prefer DNA evidence that places the scientific research, as Hacking suggests? What is the umbrella term defendant at the scene of the crime to the testimony of an eye-witness. “objectivity” actually good for beyond generating confusion and dis- Eye-witness testimony is replete with well-known epistemic risks, e.g., agreement? Are we stuck with the term simply because scientific and racial bias, conflicts of interest, the unreliable and creative nature of legal discourse is shot through with talk about objectivity? Is that suf- human memory, and outright lying; whereas hard physical evidence is ficient reason to preserve an idle wheel in our epistemic and phronetic publicly available, out there for everyone to see. The prosecutor can tell practices? the jury, “The objective evidence is right before your nose. All you have We reject Hacking's scepticism on two grounds. First, on Hacking's to do is look.” Physical evidence is objective, then, in the sense that own terms it matters normatively that objectivity is doing work in our relying on it averts, at least to some degree, epistemic risks like personal knowledge acquisition practices. Hacking, following J.L. Austin (1961), bias, conflicts of interest, fallible human memory, and dishonest testi- enjoins us to examine terms in their “sites of use” (pp. 14–15), and he mony. The very publicity of physical evidence makes it objective, al- appeals to the case of the ‘Commission of Inquiry into the Decline of though this by no means implies that such evidence can't lead us astray. Sockeye Salmon in the Fraser River’ as evidence that the concept of It's simply to say that there are reasonable grounds to prefer it to less objectivity is not used in practice. However, we argue that objectivity reliable alternatives. allows us to maintain the division of epistemic labour. That is, when we Again, as we saw in the scientific context, evidence in the law is also look to its sites of use we see that it is not merely some esoteric phi- called objective in virtue of the process by which it is acquired. For losophical concept, but rather that it forms part of a socially useful example, law consultations on matters of public policy invariably in- practice. And on Hacking's Austin-inspired account, that a term is used volve the review of relevant scientific evidence. Consider, for instance, in practice provides us with good reason to use it in philosophy. the final report from the Wessely Review – an independent reviewof Secondly, we can also find some guidance for the usefulness of the the Mental Health Act (1983) commissioned by the UK Department of abstract concept of objectivity in an analogous debate in political phi- Health – which states that, on top of a robust consultation process, “we losophy over the usefulness of the concept of ‘justice’ – this is typically spent time assessing existing evidence and data in relation to the Mental referred to as the debate between ‘ideal’ and ‘non-ideal’ theory. Health Act, and commissioned bespoke data analysis to inform our Amartya Sen (2006), for instance, argues that we do not need an ideal findings” (Department of Health and Social Care, 2018, p. 41). More- theory of justice in order to pursue actual justice in the real world, and over, the report assures us that not only was the existing evidence that the ideal theory can even be a hindrance to this practical work. By considered, but a premium was placed on evidence reached via the contrast, John Rawls (1999), argues that we need the ideal theory so most rigorous available scientific methods. And this is all meant to that we have something to aim at (pp 89–90). While it is beyond the assure the public that the evidence they relied on was objective—that it scope of this paper to engage fully in debate over ideal versus non-ideal was reached via methods that manage important epistemic risks. theories of justice (see Laura Valentini (2012) for a more complete Finally, to drive home the difference between the two types of ob- articulation of this debate), this at least shows us the appeal of abstract jective evidence we've discussed, consider an example that highlights philosophical concepts, such as justice and objectivity. Consider, for the distinction. Say that the court has a choice between E1 and E2.E1 is instance, only being able to point to individual instances of social physical evidence acquired in accordance with the proper evidence- wrongs in the world typically associated with injustice; we could, for handling procedures we discussed in §4.2; and E2 is physical evidence example, identify cases of police brutality and say that something so- identical to E1 except that it was acquired in violation of the standard cially bad has happened there, but it would still be useful to appeal to procedures of evidence management. Despite being physically iden- the abstract concept of injustice to explain that badness. Similarly, tical, the court would prefer E1 to E2, because although both avert while we could point to particular instances of scientific research vice certain epistemic risks in virtue of their publicity, E1 averts additional (as Hacking recommends), such as bias, it would still be useful to be important epistemic risks in virtue of the manner it was acquired. In able to refer to the ideal of objectivity to explain what is going wrong 16 other words, although physically identical, the court would consider E1 when these vices occur. Our commitment to justice as a normative more objective than E2. Thus, in science and law, we refer to evidence as objective due to i) its public nature and/or ii) the epistemic-risk- mitigating processes whereby we obtain it. 16 Jack Wright (2018) makes a similar point, arguing that objectivity is a ‘relational term’, by which he means that “objectivity helps to relate practices to 5. Another word from our skeptics other practices and to the abstract uses, goals, ideals, and types of practices associated with types of objectivity” (p. 389). Thinking of Wright's suggestion At this point we should return to the deflationary skeptical attacks as analogous to the debate between ideal versus non-ideal theory (as described on objectivity we mentioned at the outset, to see whether we've done above) is useful, just because of how much material that debate has produced, enough to answer them. We consider our answer to MacKinnon in §4.5 thus giving an idea of what the relational concept might look like.

68 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70 ideal keeps us vigilant in the fight against known manifestations of References injustice; but it also keeps us on the lookout for a) hitherto unknown tokens of known types of injustice and b) unknown types of injustice; Aristotle (2000). In R. Crisp (Ed.). Nicomachean ethics. Cambridge: Cambridge University and when we find these unknown tokens and types it affords usan Press. Austin, J. L. (1961). Philosophical papers. Oxford: Oxford University Press. abstract concept to appeal to as we make sense of and learn how to Biddle, J. B., & Kukla, R. (2017). The geography of epistemic risk. In K. C. Elliot, & T. manage them. Our commitment to objectivity serves a similar role our Richards (Eds.). Exploring inductive risk: Case studies of values in science (pp. 215–235). attempts to get things right, promoting vigilance in our efforts to Oxford: Oxford University Press. Brink, D. O. (2001). Legal interpretation, objectivity, and morality. In B. Leiter (Ed.). manage known epistemic and phronetic risks, keeping us on the lookout Objectivity in law and morals (pp. 12–65). Cambridge: Cambridge University Press. for new tokens and types of such risks, and helping us make sense of Buck v. Bell, 274 U.S. 200 (1927). and manage these new tokens and types as they arise. As an organising California v. Brown 479 U.S. 538 (1987). ideal for our practices, objectivity, like justice, does important work. CC v KK & STCC (2012). EWCOP. 2136. Charlton, J. I. (1998). Nothing about us without us: Disability oppression and empowerment. University of California Press. 6. Conclusion Conee, E., & Feldman, R. (2008). Evidence. In Q. Smith (Ed.). Epistemology: New essays (pp. 83–104). Oxord: Oxford University Press. Crenshaw, K., Gotanda, N., & Peller, G. (Eds.). (1995). Critical race theory: The key writings Taking inspiration from Koskinen's (2018) recent work, this paper that formed the movement. New York: The New Press. has proposed a shared rescue strategy for objectivity in science and the Daston, L., & Galison, P. (2007). Objectivity. New York: Zone Books. law. In science, we need assurance that we can rely on each other as Department of Health and Social Care (2018). Modernising the Mental Health Act. Increasing choice, reducing compulsion. Final report of the Independent Review of the epistemic agents, which requires guarding against a variety of very Mental Health Act 1983. Crown Copyright. specific epistemic risks in very context-specific ways. Objectivity isthe Douglas, H. (2004). The irreducible complexity of objectivity. Synthese, 138(3), 453–473. assurance that such epistemic risks have been managed. This is Koski- Dworkin, R. (1986). Law's empire. Cambridge: Harvard University Press. Dworkin, R. (1996). Objectivity and truth: you'd better believe it. Philosophy & Public nen's argument. We argue that the same is true in the law. The law's Affairs, 25(2), 87–139. legitimacy rests on its claim to objectivity. In other words, to be legit- Elgin, C. Z. (2017). True enough. Cambridge: The MIT Press. imate, the law needs to manage a range of epistemic, ethical and Epstein, S. (1996). Impure science: Aids, activism, and the politics of knowledge. Berkeley and Los Angeles, California: University of California Press. phronetic risks. We argue that legal objectivity is the assurance that Eskridge, W., & Ferejohn, J. (1994). Politics, interpretation, and the rule of law. Nomos, context-specific measures have been taken to manage such risks, and 36, 265–294. we have provided an indicative list of what some of the objectivity- European Court of Human Rights (2012). Guide on article 5: Right to liberty and security assuring strategies in the law currently are or should be. This list is not article 5 of the convention. ECHR Research: Division. Eysenck, H. J., Wakefield, J. A., & Friedman, A. F. (1983). Diagnosis and clinical as- exhaustive, and it will doubtlessly change as new sources of epistemic sessment: The DSM-III. Annual Review of Psychology, 34, 167–193. and phronetic risks in the law emerge. Foster, S. (2008). Should courts give stare Decisis effect to statutory interpretation We believe that this approach to objectivity in the law can move us methodology. The Georgetown Law Journal, 96(6), 1863–1991. Fricker, M. (2007). Epistemic injustice: Power and the ethics of knowing. Oxford: Oxford beyond the dialectical stalemate in the current conversation about ob- University Press. jectivity. Critics argue that genuinely determinate and objective legal General Assembly, U. N. (2006). Convention on the Rights of Persons with Disabilities: re- judgments are utterly beyond our reach, and so we should abandon all solution/adopted by the General Assembly. 24 January 2007, A/RES/61/106, available at: http://www.refworld.org/docid/45f973632.html [accessed 15 October 2018]. pretence of the law's objectivity. Defenders of objectivity insist that the Goldberg, S. (2011). The division of epistemic labour. Episteme, 8(1), 112–125. ideal is indispensable, but their insistence on conceptualizing objec- Goldman, A. I. (2005). Legal evidence. In M. P. Martin, & W. A. Edmundson (Eds.). The tivity as determinacy fails to do justice to the legitimate concerns raised Blackwell guide to the philosophy of law and legal theory (pp. 315–323). New York: John Wiley & Sons. by the critics of objectivity, and so tends to alienate them. The risk Grasswick, H. (2010). Scientific and lay communities: Earning epistemic trust through account of objectivity, we believe, can satisfy disputants on both sides knowledge sharing. Synthese, 177(3), 387–409. of this divide. For it acknowledges the power of the best critiques of Guest, S. (2007). Objectivity and value: Legal arguments and the fallibility of judges. In M. Freeman, & R. Harrison (Eds.). Law and philosophy (pp. 76–103). Oxford: Oxford objectivity; indeed, it sees these critiques as an essential part of the University Press. objective stance, as they identify the very epistemic and phronetic risks Hacking, I. (2015). Let's not talk about objectivity. In F. Padovani, A. Richardson, & J. Y. that objective legal practice must manage. But the risk account of ob- Tsou (Eds.). Objectivity in science: New perspectives from science and technology studies jectivity also does justice to the fact that the law is saturated with the (pp. 19–33). New York: Springer. Harding, S. (1991). Whose science? Whose knowledge? Thinking from women's lives. Ithaca: language of objectivity and that the ideal does important work in Cornell University Press. practice. We are hopeful, then, that this approach will help us make Harding, S. (1992). After the neutrality ideal: Science, politics, and “strong objectivity”. progress in contemporary debates about objectivity in science, law, and Social Research, 567–587. Harpur, P. (2009). Sexism and racism, why not ableism? Calling for a cultural shift in the in the domains where such concerns intersect, e.g., the intersection of approach to disability discrimination. Alternative Law Journal, 34(3), 163–167. law and psychiatry. If we listen to the voices of all concerned, and take Hobbes, T. (1651/1994). Leviathan. In E. Curley (Ed.). Leviathan, with selected variants seriously the epistemic and phronetic risks they call to our attention, a from the Latin edition of 1668. Hackett: Indianapolis. Kitcher, P. (2001). Science, truth and democracy. Oxford: Oxford University Press. workable approach to objectivity might just be within our fallible, Kitcher, P. (2011). Science in a democratic society. New York: Prometheus Books. imperfect human reach. Knight, J., & Johnson, J. (1994). Public choice and the rule of law: Rational choice theories of statutory interpretation. Nomos, 36, 244–264. Koskinen, I. (2018). Defending a risk account of scientific objectivity. The British Journal Acknowledgments for the Philosophy of Science. https://doi.org/10.1093/bjps/axy053 axy053. Leiter, B. (2001). Objectivity, morality, and adjudication. In B. Leiter (Ed.). Objectivity in Matt Burch worked on this paper while on research leave supported law and morals (pp. 66–98). Cambridge: Cambridge University Press. Leiter, B. (2002). Law and objectivity. In J. L. Coleman, S. Shapiro, & K. E. Himma (Eds.). by funds from both the Wellcome Trust (Mental Health and Justice Grant The Oxford handbook of jurisprudence and philosophy of law (pp. 969–989). Oxford: agreement No 203376/Z/16/Z) and the Independent Social Research Oxford University Press. Foundation (Early Career Fellowship). Longino, H. (1990). Science as social knowledge: Values and objectivity in scientific inquiry. Katherine Furman conducted research for this paper while receiving Princeton: Princeton University Press. MacKinnon, C. A. (1983). Feminism, Marxism, method, and the state: Toward feminist funding from the European Research Council (ERC) under the European jurisprudence. Signs: Journal of Women in Culture and Society, 8(4), 635–658. Union's Horizon 2020 research and innovation programme (Grant MacKinnon, C. (1987). Feminism unmodified: Discourses on life and law. Cambridge, Mass: agreement No 667526 K4U). Harvard University Press. MacKinnon, C. (1991). Reflections on sex equality under law. The Yale Law Journal, 100(5), 1281–1328. https://doi.org/10.2307/796693. Declaration of interest McDowell, J. (2007). What myth? Inquiry, 50, 338–351. McMullin, I. (2018). Existential flourishing: A phenomenology of the virtues. Cambridge: Cambridge University Press. The authors have no conflicts of interest to declare.

69 M. Burch and K. Furman International Journal of Law and Psychiatry 64 (2019) 60–70

Miller, D. (2017). Justice. In E. N. Zalta (Ed.). The Stanford Encyclopedia of Philosophy (Fall Sen, A. (2006). What do we want from a theory of justice? Journal of Philosophy, 103(5), 2017 Edition), URL = https://plato.stanford.edu/archives/fall2017/entries/justice/. 215–238. Montuschi, E. (2016). Using science, making policy: What should we worry about? Shafer, L. R. (2003). Moral realism: A defence. Oxford: Oxford University Press. European Journal of Philosophy of Science, 7(1), 57–78. Shah, P., & Mountain, D. (2007). The medical model is dead–long live the medical model. Munro, E., & Hardie, J. Why we should stop talking about objectivity and subjectivity in The British Journal of Psychiatry, 191(5), 375–377. social work. British Journal of Social Work. (2018) Sibley, F. (2001). Approach to aesthetics: Collected papers on philosophical aesthetics. Oxford: Nagel, T. (1986). The view from nowhere. New York: Oxford University Press. Oxford University Press. Nussbaum, M. (2001). Political objectivity. New Literary History, 32(4), 883–906. Stegenga, J. (2018). Medical Nihilism. Oxford: Oxford University Press. Parascandola, M. (2010). Epistemic risk: Empirical science and the fear of being wrong. Susskind, R. E. (Ed.). (2005). The Susskind interviews: Legal experts in changing times. Law, Probability and Risk, 9, 201–214. London: Sweet & Maxwell. Pies, R. (2007). How “objective” are psychiatric diagnoses?:(guess again). Psychiatry Szasz, T. S. (1974). The myth of mental illness: Foundations of a theory of personal conduct (Edgmont), 4(10), 18–22. (Rev. ed.). New York: Harper & Row. Postema, G. (2001). Objectivity fit for law. In B. Leiter (Ed.). Objectivity in law and morals UN Committee on the Rights of Persons with Disabilities (2014). General comment no 1, (pp. 99–143). Cambridge: Cambridge University Press. article 12: Equal recognition before the law. CRPD/C/GC/1https://documents-dds- Rawls, J. (1999). The law of the peoples. Cambridge, Massachusetts: Harvard University ny.un.org/doc/UNDOC/GEN/G14/031/20/PDF/G1403120.pdf?OpenElement. Press. Unger, R. M. (1986). The critical legal studies movement. Cambridge, Massachusetts: Reiss, J., & Kitcher, P. (2009). Biomedical research, neglected diseases, and well-ordered Harvard University Press. sciences. Theoria, 24(66), 263–282. Wakefield, J. C. (1992). The concept of mental disorder: On the boundary between bio- Reiss, J., & Sprenger, J. (2017). Scientific objectivity. In E. N. Zalta (Ed.). The Stanford logical facts and social values. American Psychologist, 47(3), 373–388. Encyclopedia of Philosophy. Winter 2017 Edition), URL = https://plato.stanford.edu/ Waldron, J. (2011). The rule of law and the importance of procedure. Nomos, 50, 3–31. archives/win2017/entries/scientific-objectivity/. Williams, B. (1985). Ethics and the limits of philosophy. London and New York: Routledge. Resnik, D. B. (1998). Conflicts of interest in science. Perspectives on Science, 6(4), Wolbring, G. (2008). The politics of ableism. Development, 51(2), 252–258. 381–408. Wright, J. (2018). Rescuing objectivity: A Contextualist proposal. Philosophy of the Social Rudner, R. (1953). The scientist qua scientist makes value judgments. Philosophy of Sciences. 48(4), 385–406. Science, 20(1), 1–6. Wylie, A. (2003). Why standpoint matters. In R. Figueroa, & S. Harding (Eds.). Science and Schemann, N. (2011). Shifting ground: Knowledge and reality. Oxford: Oxford University other cultures: Issues in philosophies of science and technology (pp. 26–48). New York: Press. Routledge.

70